Informed Consent in an International Trial of Stress Ulcer Prophylaxis - Patterns and Predictors: A Protocol and Statistical Analysis Plan

BackgroundInformed consent rates are inconsistently incorporated in trial reports, and literature on consent patterns and predictions is sparse, particularly in the field of critical care. ObjectiveThe overall objective of this study is to describe the patterns and predictors of consent rates in REVISE. The specific aims are to analyze the consent models used, consent rates, participants in the consent encounter, reasons for declined consent, and factors associated with obtaining consent. MethodsThis is a pre-planned secondary study of the REVISE Trial (NCT03374800) which compared pantoprazole to placebo on the outcome of clinically important upper gastrointestinal bleeding among invasively ventilated patients in the intensive care unit (ICU). Research ethics committees approved the protocol in all jurisdictions. Research personnel prospectively collected standardized data for each consent encounter, including the consent model (a priori, deferred, or opt-out), the role of the individual who provided or declined consent (patient, SDM, other), and the reasons for declined consent, the role of the individual who requested consent (research coordinator, site investigator, ICU physician) and the consent encounter method (in person or via telephone). When consent was provided and then later revoked, who revoked consent (patient, SDM, other) and timing (in ICU, in hospital, post hospital), as well as details about permission for data retention were collected. In this study, consent rates will be calculated across REVISE sites, and in relation to the COVID-19 pandemic. We will also calculate the consent rates for a priori and deferred consent models, the timing from deferred recruitment to consent provided or declined, which personnel requested consent (research coordinator, site investigator, ICU physicians, other), and who engaged in the consent encounter for each consent model (patient, SDM, other). Multilevel logistic regression analysis will be conducted to evaluate variables independently associated with consent including additional site and staff variables. ResultsBy analyzing the frequency and impact of consent models, and consent encounters with various stakeholders, results will highlight the acceptability of different approaches, and the impact of different approaches in this critical care trial. ConclusionsThis pre-planned retrospective sub-study using an international randomized controlled trial database will provide useful informed consent metrics and knowledge that is relevant to contemporary global trial conduct.